Customers’ Multihoming Behavior in Ride-Hailing: Empirical Evidence from Uber and Lyft

Chitla, Sandeep and Jagabathula, Srikanth and Mitrofanov, Dmitry and Cohen, Maxime C. Available at SSRN: https://ssrn.com/abstract=4591826 or http://dx.doi.org/10.2139/ssrn.4591826

The author collected data from 1.4M rides, 162k riders, NYC, 2018 through aelf-selected third-party panel. Uber consists of 65.7% of rides. Over a full year, 56.7% of riders used only Uber, 26.7% only Lyft, 16.6% used both at least once. The paper used subsample of the multihoming riders to estimate a structural choice model with Bayesian price/wait-time learning and a search-cost-driven consideration stage. The finding: only 16.6% of riders ever used both platforms, and even they checked both apps just 43.4% of the time — inertia and brand affinity, not price or wait time, dominate the choice.

Personally I don’t think it’s really inertia and brand affinity. When price are similar most of the time, what’s the point?